1. Field
This disclosure is directed to microfluidic devices and methods for diagnostic, molecular, and biochemical assays and, more particularly, to microfluidic technologies for dispensing and distributing fluid from on-cartridge reagent reservoirs, for pumping, heating and mixing, and for rehydrating dried reagents without bubble entrainment and without reagent washout.
2. Description of Related Art
Microfluidic devices have found increasing use as tools for diagnostic assays. The devices described by Wilding in U.S. Pat. No. 5,304,487 consisted of “mesoscale” channels and chambers formed on reusable silicon substrates which were infused with fluid reagents from off-cartridge syringe pumps. No consideration was given to on-cartridge fluid and reagent storage and delivery. However, practical commercial applications have lead in the direction of “consumable” cartridges—disposable, single use “sample-to-answer” cartridges that are self-contained for all reagents needed for a particular assay or panel of assays. This is particularly true in the case of molecular biological assay applications, where contamination associated with sample carryover or handling absolutely must be avoided.
On board reagents may include both liquid and dry reagent forms. Both such reagent classes have been subject to certain problems in realization of successful products. Here we address liquid handling issues associated with initial wetout of the channels and chambers of the cartridge and with rehydration of dried reagents. During filling and operation of a cartridge containing microfluidic channels and chambers, particularly those cartridges having a plastic body, liquid wetout is often uneven, such that air pockets are not infrequently entrained in the fluid column by the advancing meniscus against surfaces and in corners. During pumping and mixing of biological samples, foam and bubbles may form that negatively impact the assay performance of the device. Bubbles may arise due to uneven filling of channels or chambers containing dried reagents. Reagent rehydration, wetout and venting are interlinked with the problem of bubble formation. The problem is exacerbated in more complex fluid networks such as described in U.S. Pat. Nos. 6,068,752 to Dubrow and 6,086,740 to Kennedy, for example, and in capillary flow-driven devices such as described by Buechler in US Patent Application No. 2005/0136552 or Wyzgol in US Patent Application No. 2004/024051, which have proved notoriously difficult in plastic body devices.
Bubbles may also arise during heating of a sample liquid due to degassing. It is well known that gas solubility is inversely related to temperature and that solutions which are heated readily become supersaturated. Also a source of bubbles by degassing is cavitation, where a fluid is sheared, such as during mechanical or ultrasonic mixing in microfluidic cavities.
Bubbles interfere with optical interrogation of liquids in microfluidic “cuvettes”. The path of light may be altered due to a lensing effect created by the curvature of the gas bubble surface and/or due to the gas bubble refracting the light. Bubbles may also interfere with biochemical reactions by altering solute concentrations at bubble interfaces, by denaturing protein structure, and by impacting bulk heating rate and the homogeneity of temperature in a liquid. For example, in the PCR reaction, in which a thermostable polymerase is used to amplify copies of a target nucleic acid, heating and cooling is uneven in the presence of bubbles in the fluid, reducing the efficiency of the process and limiting sensitivity. The presence of bubbles also reduces the volume of fluid in the reaction chambers, and in assays which rely on detecting analyte in volumes of 10-50 uL or less, the presence of a large trapped bubble in a reaction chamber can effectively kill the assay.
In reactions that rely on rate determination, bubbles can drastically interfere with optical determination of slopes and with homogeneous rapid rehydration of dried reagents as is needed to start the reaction with proper availability of substrates. A variety of dried reagents, such as a fluorescent probe, enzyme, buffer or control analyte, may be placed within chambers of a microfluidic device and are needed for proper conduct of the assay. During wetout, entrapment of one or more bubbles may result in incomplete dissolution and mixing of the dry reagent and the sample, thereby impairing the reaction efficiency and reducing the sensitivity of the test.
Lei, in U.S. Pat. No. 6,637,463 proposes varying flow impedance in parallel channels through use of surface tension features and/or cross-sectional area so as to equalize pressure drops, and hence flow, through the multiple flow paths. In one instance, a plurality of exit channels is used to drain fluid from a well so as to avoid formation of recirculating currents or fluid stagnation that would otherwise tend to inefficient washing of fluid and trapping of air bubbles. However, each such feature must be designed by trial and error, and the designs are thus not robust or readily adapted for different assays. Because microscopic variations in dimensions and surface chemistry are difficult to control in microfluidic circuit manufacture, the methods have not been proven a practical solution to the problem of equally dividing flow between parallel subcircuits within a microfluidic card. No description of the use of diaphragms with features for improving wetout was offered.
Ulmanella (US Patent Application No. 2007/0280856) reported efforts to control the meniscus of a fluid filling a microfluidic chamber by physically modifying the bottom surface of the chamber, for example by installing an energy barrier to slow down or stop the leading edge of the meniscus as it crosses the floor of the chamber, or by use of a plurality of grooves or posts on the bottom surface, or by sculpting the depth of the chamber so as to modulate capillary action, or by using a syringe pump, by centrifugation, or by application of a vacuum on the outlet side of the chamber. None of these methods has proved a practical solution to the problem. Capillary action is highly unpredictable and tends to promote formation of air pockets and use of a syringe pump or application of vacuum, as commonly practiced in the prior art, tends to shear the fluid and drive fluid down the path of least resistance, further exacerbating the problem. For example, when two or more microfluidic channels branching from a single inlet are presented to a fluid, such as is useful for splitting a sample or reagent between multiple diagnostic assays pathways in parallel, the fluid may fill the path most readily wetted and leave empty the path having higher fluid resistance. Very tiny differences in resistance between channels lead to preferential wetting of a single channel and no wetting of branching parallel channels, a problem well known to those skilled in the art.
Ulmanella further addresses the effect of dried reagents in wetout of microfluidic chambers and concludes that filling efficiency of chambers containing center-spotted dried reagent was less than 50%, chambers having inlet side spotted reagent were wetted at 65% efficiency, but for chambers having outlet side spotted reagent, the filling efficiency without bubbles increased to 95%. However, positioning of reagent spots with millimeter accuracy during manufacturing is neither a necessary nor a satisfactory means of achieving wetout in the presence of dried reagent spots because it is preferential that the chamber be fully wetted before the reagent is rehydrated so that the concentration of the reagent is not diluted by washout into a downstream channel, as is highly likely if the dry reagent is positioned at the downstream outlet from the chamber!
It is further known that reduction in interfacial and surface tensions in the microfluidic channels or chambers can be achieved, for example, by plasma treatment of the substrate(s) or incorporation of surfactants to decrease hydrophobicity, and by applying a radius to channel intersections. These treatments are also known to improve wettability, but are not effective in eliminating mechanically entrained bubbles and bubbles resulting from thermal degassing, cavitation or stagnation zones. In fact, surfactants can increase the propensity of the gaseous phase to form stable bubbles and foams which can defeat performance of the assay by their persistence. Moreover, the modification of surfaces by processes such as plasma treatment are anticipated to be difficult to control in manufacturing and may be impermanent, degrading progressively during device storage. Therefore it is desirable and is an object of this invention to develop mechanical means and methods for reducing the formation and entrainment of bubbles during initial wetout of assay channels, during rehydration of dry reagents, and for preventing or reducing accumulation and interference of bubbles during operation of the device.